As an in-vehicle display device, there is a monocular head-up display (HUD) by which driving information such as a vehicle velocity, a movement direction, etc. can be recognized visually.
In such an HUD, a viewer's eye position is derived from a photographic image of a viewer's head. The angle and position of a flat mirror are controlled automatically based on a result of the derivation. Video is presented to one eye of a viewer following the motion of the viewer's head.
In such display device, video cannot be presented to one eye of a viewer robustly because the influence of external light or the like may make it difficult to derive the position of the viewer's eye from a photographic image.